On-line identification based optimal control method

Abstract

This paper presents an On-line Identification based Optimal Control (OIOC) method that composed of two layers: the control layer is used to realize the close-loop control while the optimization layer is used to search the optimal control parameters. This control strategy can ensure the robustness of a servo system. The nonlinear PID controller, Elman network based identifier and Particle Swarm Optimization (PSO) algorithm are designed and applied to OIOC. The simulation results with respect to a typical electro-hydraulic servo system are presented.